There are two approaches to solving biological questions at the molecular level: structural analysis and functional analysis. Both approaches make use of independent technologies that have been optimized to provide particular types of information. Traditionally, the principal techniques employed for structural analysis are microscopy (optical, electron, and force), optical spectroscopy (infrared, visible and ultraviolet), nuclear magnetic resonance spectrometry, mass spectrometry, x-ray crystallography, sequencing, and molecular modeling. These techniques provide information about the size of the molecule in question, the kinds of functional groups that are present in the molecule, and the relative orientation of those functional groups in one, two and three dimensions.
Although researchers have been able to collect a wealth of structural information about a wide diversity of molecules using the techniques mentioned above, the relationship between structure and function is still largely obscure. Most biologically interesting molecules fulfill their natural function through interaction(s) with other molecules. Thus, analytical techniques used for the study of the interactions between and among biologically interesting molecules will significantly advance the understanding of their function(s), as well as begin to more completely connect structural information with functionality. The interactions of interest can be described in terms of rate constants (association, dissociation, mass transport, diffusion and the like), equilibrium constants (measures of the affinities of the various molecules for each other) and the stoichiometry of the interaction. The ability to accurately measure and compare these fundamental parameters is critical to the characterization and understanding of molecular interactions.
Bioconjugation is a descriptive term for the joining of two or more different molecular species by chemical or biological means, in which at least one of the molecular species is preferably a biological macromolecule. These biological macromolecules include, but are not limited to, conjugation of proteins, peptides, polysaccharides, hormones, nucleic acids, lipid bilayers, liposomes and cells, with each other or with any other molecular species that add useful properties, including, but not limited to, drugs, radionuclides, toxins, haptens, inhibitors, chromophores, fluorophores, ligands, and the like. Immobilization of biological macromolecules is also considered a special case of bioconjugation in which the macromolecule is conjugated, either reversibly or irreversibly, to an insoluble support, such as a chromatographic matrix, microwell plate, porous membrane, polymer bead, glass microscope slide, silicon chip, and the like. Bioconjugation is utilized extensively in biochemical, immunochemical and molecular biological research. Major applications of bioconjugation include, but are not limited to, detection of gene probes, enzyme-linked immunological solid-phase assays, affinity purification, monoclonal antibody-drug targeting and medical imaging.
Bioconjugates are generally classed either as direct or indirect conjugates. Direct conjugates encompass those in which two or more components are joined by direct covalent chemical linkages. Alternatively, indirect conjugates encompass those in which two or more components are joined via an intermediary complex involving a biological macromolecule.
Although numerous methods of indirect bioconjugate preparation have been described, a significant number of those reported in the literature have been prepared by exploiting the biotin-avidin interaction. In this system, the binding specificity of the protein avidin (purified from egg white) or streptavidin (purified from the bacterium Streptomyces avidinii) toward the cofactor biotin (vitamin H) is utilized to connect a (strept)avidin-conjugated macromolecule with a biotin-conjugated macromolecule. Both avidin and streptavidin possess four biotin binding sites of very high affinity (Kd=10−15 mol−1).
The avidin-biotin system has been utilized extensively for enzyme-linked immunological solid-phase assays (ELISA), in which an enzyme-avidin conjugate (useful for sensitive detection by reaction with a substrate of the enzyme to produce a colored or luminescent product) is employed to detect the presence of biotin-conjugated antibody, after first binding that antibody conjugate to an immobilized hapten or antigen. Applications of the avidin-biotin system number in the hundreds, and have recently been reviewed (see, Wilchek, M. and Bayer, E. A. (1990) Methods in Enzymology, volume 184).
While utilized extensively, several limitations are known to be associated with avidin-biotin system. These include nonspecific binding of assay components due predominantly to the basicity of the avidin molecule or to the presence of carbohydrate chains on the avidin molecule, and background interference associated with the presence of endogenous biotin, which is ubiquitous in both eukaryotic and prokaryotic cells.
An alternative indirect bioconjugation system designed to circumvent the limitations of the avidin-biotin system has been recently developed for the detection of gene probes using and enzyme-linked approach (see, Kessler, C., Hôltke, H.-J., Seibl, R., Burg, J. and Mühlegger, K. (1990) Biol. Chem. Hoppe-Seyler 371, 917–965). This system involves the use of the steroid hapten digoxigenin, an alkaloid occurring exclusively in digitalis plants, and the Fab fragments derived from polyclonal sheep antibodies directed against digoxigenin (α-digoxigenin). The high specificity of the various a-digoxigenin antibodies for digoxigenin affords low backgrounds and eliminates the nonspecific binding often observed in the avidin-biotin system. Digoxigenin-conjugated DNA and RNA probes can detect single-copy sequences in human genomic Southern blots. The development of the digoxigenin-α-digoxigenin system has been reviewed (see, Kessler, C. (1990) in Advances in Mutagenesis Research (Obe, G., ed.) pp. 105–152, Springer-Verlag, Berlin and Heidelberg). The digoxigenin-α-digoxigenin system is the most recent representative of several hapten-antibody-based indirect conjugation systems now routinely used for bioconjugation.
Phenylboronic acids are known to interact with a wide range of polar organic molecules having certain requisite functionalities. Complexes of varying stabilities involving 1,2-diols, 1,3-diols, 1,2-hydroxy acids, 1,3-hydroxy acids, 1,2-hydroxylamines, 1,3-hydroxylamines, 1,2-diketones and 1,3-diketones are known to form with either neutral phenylboronic acid or phenylboronate anion. Consequently, immobilized phenylboronic acids have been exploited as chromatographic supports to selectively retain, from complex biological samples, those molecular species having the requisite functionalities. Many important biological molecules including carbohydrates, catecholamines, prostaglandins, ribonucleosides and steroids contain such functionalities, and have been either analyzed or purified in this manner. The use of phenylboronic acid chromatographic media for the separation and isolation of biological molecules has been discussed in several reviews (see, Singbal, R. P. and DeSilva, S. S. M. (1989) Adv. Chromatog. 31, 293–355; Mazzeo, J. R. and Krull, I. S. (1989) BioChromatog. 4, 124–130; and Bergold, A. and Scouten, W. H. (1983) in Solid Phase Biochemistry (Scouten, W. H., ed.) pp. 149–187, John Wiley and Sons, New York).
Phenylboronic acid, like boric acid, is a Lewis acid, and ionizes not by direct protonation, but rather by hydration to give the phenylboronate anion (pKa=8.86). Phenylboronic acid is three times as strong an acid as boric acid. Ionization of phenylboronic acid is an important factor in complex formation, in that, upon ionization, the boron nucleus changes from trigonal coordination (having average bond angles of 120° and average bond lengths of 1.37 angstroms) to tetrahedral coordination (having average bond angles of 109° and average bond lengths of 1.48 angstroms).
Molecular species having cis or coaxial 1,2- or 1,3-diol functionalities (notably carbohydrates) are known to complex with immobilized phenylboronate anions, forming cyclic esters, under alkaline aqueous conditions (see, Lorand, J. P. and Edwards, J. O. (1959) J. Org. Chem. 24, 769). Acidification of 1,2- and 1,3-diol complexes with phenylboronic acid to neutral pH is known to release the diol-containing species, presumably due to hydrolysis of the cyclic ester. Coplanar aromatic 1,3-diols, such as 1,8-dihydroxynaphthalene, are known to form stable complexes even under acidic conditions, due to the hydrolytic stability of six-membered cyclic boronic acid esters (see, Sienkiewicz, P. A. and Roberts, D. C. (1980), J. Inorg. Nucl. Chem. 42, 1559–1571).
Molecular species having pendant 1,2-hydroxylamine, 1,3-hydroxylamine, 1,2-hydroxyamide, 1,3-hydroxyamide, 1,2-hydroxyoxime and 1,3-hydroxyoxime functionalities are also known to reversibly complex with phenylboronic acid under alkaline aqueous conditions similar to those associated with the retention of diol-containing species (see, Tanner, D. W. and Bruice, T. C. (1967) J. Amer. Chem. Soc. 89, 6954).
2-Acetamidophenylboronic acids have been proposed as potential linkers for selective bioconjugation via the vicinal diol moieties of the carbohydrate residues associated with glycoproteins (see, Cai, S. X. and Keana, J. F. W. (1991) Bioconjugate Chem. 2, 317–322). Phenylboronic acid bioconjugates derived from 3-isothiocyanatophenylboronic acid have been successfully utilized for appending radioactive technetium dioxime complexes to monoclonal antibodies for use in medical imaging (see, Linder, K. E., Wen, M. D., Nowotnik, D. P., Malley, M. F., Gougoutas, J. Z., Nunn, A. D., and Eckelman, W. C. (1991) Bioconjugate Chem. 2, 160–170; Linder, K. E., Wen, M. D., Nowotnik, D. P., Ramalingam, K., Sharkey, R. M., Yost, F., Narra, R. K. and Eckelman, W. C. (1991) Bioconjugate Chem. 2, 407–414).
3-Aminophenylboronic acid has been covalently appended to proteins by a variety of chemical methods and the resulting phenylboronic acid bioconjugates tested for their binding of D-sorbitol, D-mannose and glycated hemoglobin (GHb). The interactions proved to be reversible and of very low affinity, rendering the bioconjugates of very limited practical use. Similarly, an alkaline phosphatase-phenylboronate used in an attempted enzyme-linked assay for the detection of GHb failed to detect the presence of glycated protein (see, Frantzen, F., Grimsrud, K., Heggli, D. and Sundrehagen, E. (1995) Journal of Chromatography B 670, 37–45).
A novel class of phenylboronic acid reagents and boronic acid compound complexing reagents have been developed for conjugating biologically active species (BAS) (also known as “bioactive species”) and exploiting indirect bioconjugation through reversible formation of a boronic acid complex. These reagents and associated conjugates can be used in a manner analogous to the avidin-biotin and digoxigenin-α-digoxigenin systems. However, unlike either of these two biological systems wherein the functional viability of the biological macromolecule (protein) must be maintained to preserve the requisite binding properties, the bioconjugate formed through the boron complex is generally insensitive to significant variations in ionic strength, temperature, the presence of organic co-solvents, and the presence of chaotropic agents (protein and nucleic acid denaturants). Additionally, the complex between the boronic acid and the boronic compound complexing reagent is facilely reversible, using a combination of low or high pH, elevated temperature, and/or competitive releasing reagents. These phenylboronic acid reagents and boronic compound complexing reagents, their bioconjugates and complexes as well as methods for their preparation and use are the subject of U.S. Pat. Nos. 5,594,111, 5,623,055, 5,668,258, 5,648,470, 5,594,151, 5,623,055, 5,668,257, 5,668,258, 5,688,928, 5,677,431, 5,744,727, 5,777,148, 5,837,878, 5,847,192, 5,852,178, 5,859,210, 5,869,623, 5,872,224, 5,876,938, 5,877,297, 6,008,406, 6,075,126, 6,124,471 and 6,156,884, the teachings of each of which are incorporated herein by reference.
Phosphoramidite reagents containing protected phenylboronic acid and 1,3-phenyldiboronic acid moieties for the preparation of modified synthetic oligonucleotides are the subject of U.S. Pat. No. 6,031,117. Oligonucleotides and polynucleotides containing phenylboronic acid and 1,3-phenyldiboronic acid, prepared from the aforementioned phosphoramidite reagents, are the subject of U.S. Pat. No. 6,013,783. 2′-Deoxyuridine triphosphate and uridine triphosphate containing phenylboronic acid and 1,3-phenyldiboronic acid moieties for the preparation of modified polydeoxyribonucleotides (DNA) and polyribonucleotides (RNA) are the subject of U.S. Pat. No. 5,831,046. Polydeoxyribonucleotides (DNA) and polyribonucleotides (RNA) containing phenylboronic acid and 1,3-phenyldiboronic acid, prepared from the aforementioned 2′-deoxyuridine triphosphate and uridine triphosphate reagents, are the subject of U.S. Pat. No. 5,831,045. The teachings of each these patents are incorporated herein by reference.
In recent years, an increasing number of techniques have become available for functional studies at the molecular level, enabling examination of the way(s) molecules interact to carry out their specific biological purpose(s). One of the techniques that has made a major contribution in this regard is real-time biomolecular interaction analysis, also termed BIA. BIA measures interactions between two or more molecules without the use of labels. Molecules that have been studied using this technique include proteins, peptides, nucleic acids, oligonucleotides, carbohydrates, lipids, small molecule metabolites and pharmaceuticals. Additionally, BIA has been used to study the binding of biomolecules to cell surfaces. A recent review of BIA technologies and commercial instrumentation has been published (see, Baird, C. L. and Myszka, D. G. (2001) J. Mol. Recognit. 14, 261–268), as has a recent review of BIA applications (Rich, R. L. and Myszka, D. G. (2001) J. Mol. Recognit. 14, 273–294).
One commonly used technology for performing BIA relies on the surface optical phenomenon called surface plasmon resonance (SPR), which detects minute changes in the refractive index of a medium in contact with a sensor surface. Surface plasmon resonance is the oscillation of free electrons at the surface of a thin metal film. These oscillations are affected by changes in the refractive index of the medium very near to the surface of the film. The general method for use of SPR in BIA is to immobilize one member of a biological binding pair (the ligand) on the surface of the metal film, and then to introduce the other member of the binding pair (the analyte) in solution. As the interaction progresses, mass accumulates on the sensor surface due to binding of the analyte to the ligand. This accumulation of mass is accompanied by a proportional change in refractive index at the metal surface, which is monitored by SPR in real time. The rate at which mass is gained or lost from the surface provides important functional information about the interaction. Various types of instrumentation for the implementation of SPR as a biosensor detection technique have been described (for examples, see, Liedberg, B. et al. (1983) Sensors and Actuators 4, 299–304; U.S. Pat. No. 6,127,183; U.S. Pat. No. 5,965,456; U.S. Pat. No. 5,374,563; U.S. Pat. No. 5,770,462; U.S. Pat. No. 5,064,619; U.S. Pat. No. 5,815,278; and U.S. Pat. No. 5,912,456).
An essential part of BIA using SPR detection is the means by which the ligand is tethered to the metal film surface. Common metals for SPR detection include gold, silver and copper, as well as others; the most common metal is gold, which is usually deposited in a thin layer on a material having a high index of refraction, such as glass. A common and extremely useful technique for the chemical modification of gold surfaces for the immobilization of biomolecules is the use of ω-functionalized alkanethiols to form self-assembled monolayers presenting reactive sites for the attachment of biological ligands. Several methods have been described. In one (see, Lofas, S. et al. (1990) J. Chem. Soc., Chem. Commun., 1526–1528), long-chain hydroxyalkanethiols are used to form a monolayer of exposed hydroxyl groups on a gold biosensor surface; carboxymethyldextran polymers are then covalently attached to the hydroxyl groups to provide a matrix in which ligands may be chemically immobilized. In another (see, Shumaker-Parry, J. S. et al. (2000) Proceedings of the SPIE 3922, 158–166; Spinke et al. (1993) J. Chem. Phys. 99, 7012–7019), a binary mixed monolayer composed of hydroxyl-terminated and biotin-terminated alkanethiols is deposited on a gold biosensor surface; streptavidin is next bound to the exposed biotin moieties to provide an intermediary proteinaceous layer; and finally, biotinylated ligands are contacted with the streptavidin-modified surface and immobilized. In yet another method (U.S. Pat. No. 6,197,515), a self-assembled monolayer of alkanethiols terminally modified to provide a metal chelating functionality for the coordination of a metal ion is prepared on the gold biosensor surface; ligands having an accessible, partially coordinated metal ion are then immobilized following contact with the chelating monolayer surface.
The above methods for chemically modifying a gold surface and subsequently using the surface for the immobilization of a ligand suffer from several inadequacies when applied to BIA using SPR detection. Use of covalent immobilization of the ligand within a dextran matrix can frequently compromise the biological viability of the ligand, rendering it unsuitable for highly accurate interaction analysis. Additionally, this method distributes the ligand at varying distances from the metal surface throughout the thickness of the matrix layer; since SPR detection sensitivity decreases rapidly as the distance from the surface increases, this approach seriously reduces the observed detection sensitivity of the analysis, requiring more ligand to be immobilized to observe adequate SPR responses. This method also requires that the analyte penetrate the dextran matrix in order to find its immobilized biological binding partner; the kinetics of this process can often severely complicate the overall interaction analysis. Finally, the use of covalent chemistry renders the immobilization process irreversible. The use of the biotin-streptavidin system for immobilization as described in the previous paragraph suffers from the problems of functional irreversibility due to the very high dissociation constant of the biotin-streptavidin complex, the complicating effects of endogenous biotin in biological samples, and the need to covalently label the ligand with biotin and subsequently purify the biotinylated ligand prior to immobilization. The metal chelate technique requires that an appropriate metal chelating functionality be introduced into the ligand; for genetically engineered proteins, this can be accomplished using known recombinant DNA techniques to insert a stretch of multiple histidine amino acids at a desired location in the polypeptide chain, but for non-protein ligands, incorporation of a metal coordinating functionality into the ligand is considerably challenging.
In view of the foregoing, it would be advantageous to develop an immobilization chemistry for SPR-based BIA that exploits the benefits and advantages associated with the aforementioned class of phenylboronic acid reagents and boronic compound complexing reagents, which have been developed for conjugating biologically active species and for exploiting indirect bioconjugation through reversible formation of a boronic acid complex. The present invention fulfills this and other needs.